International Journal of

Environmental Research

and Public Health

Article

Suppression of Rice Blast by Bacterial Strains Isolatedfrom Cultivated Soda Saline-Sodic Soils

Yi Wei 1,†, Lanhui Li 1,†, Wenjun Hu 1, Huiyan Ju 1, Mingzhe Zhang 1, Qingming Qin 1 ,Shihong Zhang 1 and Guihua Li 1,2,*

1 College of Plant Sciences, Jilin University, Changchun 130062, China; wyziyu@163.com (Y.W.);lilh0205@163.com (L.L.); 18264901015@163.com (W.H.); jhuiyan@163.com (H.J.); mzzhang@jlu.edu.cn (M.Z.);qmqin@jlu.edu.cn (Q.Q.); zhang_sh@jlu.edu.cn (S.Z.)

2 Department of Plant Protection, Shenyang Agricultural University, Shenyang 110866, China* Correspondence: liguihua@jlu.edu.cn† These authors contributed equally to this work.

Received: 16 June 2020; Accepted: 19 July 2020; Published: 21 July 2020�����������������

Abstract: Rice blast caused by Magnaporthe oryzae is one of the most serious rice diseases worldwide.Biological control is gaining popularity as a promising method for the control of this disease;however, more effective microbial strains with strong adaptability in rice fields need to be identified.Here, we report for the first time the successful identification of biocontrol bacterial strains fromfrozen soils of the soda saline-sodic land. We isolated 82 bacterial strains from rice fields in the westernSongnen Plain of China, one of the three major soda saline soils in the world. Five of the isolated strainsexhibited strong inhibition to M. oryzae growth. The potential strains were identified as Bacillus safensisJLS5, Pseudomonas koreensis JLS8, Pseudomonas saponiphila JLS10, Stenotrophomonas rhizophila JLS11 andBacillus tequilensis JLS12, respectively, by 16s RNA gene sequence analysis. The antagonistic assay andthe artificial inoculation tests showed that JLS5 and JLS12 could effectively inhibit conidial germinationand pathogenicity of the rice blast fungus, both preventively and curatively. The suppression ofpathogenicity was further confirmed by greenhouse experiments, showing the effectiveness of JLS5and JLS12 as a potential biological control agents of M. oryzae. The potential application of thesecold-tolerant strains for rice blast control in cold regions is discussed. Our data suggest that sodasaline-sodic soils are a rich source for biocontrol strain isolation.

Keywords: Magnaporthe oryzae; rice blast disease; biological control; saline-sodic land; soil bacteria;antagonistic bacterial strain

1. Introduction

Rice blast, caused by Magnaporthe oryzae, is the most serious disease of cultivated rice that resultsin 10–30% annual rice yield losses [1]. The blast fungus uses appressorium, a dome-shaped celldeveloped from germinated conidia, to mechanically breach the outer plant surface and infect hostcells [2]. The main disease control method currently used is the application of chemical fungicides,which, albeit effective, pose severe threats to human health and environment as well as leadingto drug resistance in fungi. Another common management strategy is to breed blast-resistant ricevarieties. However, this method has proven to be ineffective for long-term control of the disease underfield conditions [3]. Therefore, alternative disease management strategies that are both effective andenvironmentally friendly are highly desired.

Biological control of rice diseases using microorganisms has attracted increased attention inrecent years owing to its superior features such as inhibition specificity, low toxicity, and lackof pathogen resistance [4–6]. Quite a few bacterial strains isolated from various environmental

Int. J. Environ. Res. Public Health 2020, 17, 5248; doi:10.3390/ijerph17145248 www.mdpi.com/journal/ijerph

Int. J. Environ. Res. Public Health 2020, 17, 5248 2 of 12

niches have shown a positive effect toward controlling M. oryzae. Some of the examples includethe plant endophytic Bacillus tequilensis GYLH001 isolated from Angelica dahurica [7], and Bacillussafensis B21 from Osmanthus fragrans [8]. There are also bacteria isolated from the rhizosphere,including rhizobacteria [9], Bacillus cereus HS24 [10,11], and Streptomyces spp. [12] from the ricerhizosphere, and Paenibacillus terrae NK3-4 from the soybean rhizosphere [13]. In addition, some bacteriahave been isolated from other environmental niches such as Staphylococcus sp. strain LZ16 fromseawater [14], and Streptomyces globisporus JK-1 from a contaminated fungal culture [15].

Soils, especially those from saline-sodic land, are rich sources of bacteria that might exhibitantifungal effects. In this study, we isolated 82 bacterial strains from the rice fields in the westernSongnen Plain of China, one of the three major soda saline soils in the world. Nine strains showedsuppression ability against M. oryzae growth, five of which exhibited a strong inhibition effect. We furtherinvestigated in detail the antifungal effects of two strains, Bacillus safensis JLS5 and Bacillus tequilensisJLS12. Both strains could inhibit mycelial growth, conidial germination, and pathogenicity of M. oryzaeeffectively, suggesting their potential application as biological control agents against M. oryzae.This study showed that soda saline-sodic soils could be rich sources for the isolation of biocontrol strains.

2. Materials and Methods

2.1. Fungal Strains and Culture Conditions

Two Chinese M. oryzae field strains, Y34 [16,17] and JL0910 [18], were used in this study, which werecultured and maintained as reported previously [17].

2.2. Soil Sample Collection

Seven soil samples were collected from soda saline-alkali rice fields in the western Songnen Plain,located in Qian Gorlos Mongol Autonomous County, Songyuan, Jilin province of China. In brief,the surface-frozen soils were collected in December, and the samples were stored at 4 ◦C for ten daysbefore the bacterial strain was isolated.

2.3. Isolation of Bacteria from Soil Samples

Five grams of soil in 50 mL sterile water was diluted to10−6, based on ten-fold gradient dilution.An aliquot of 100 µL each dilution was spread on Luria-Bertani (LB) plates and incubated at 28 ◦C for24 h. Morphologically distinct colonies were purified and stored.

2.4. Screening of Antagonistic Strains against M. oryzae

The candidate biocontrol bacterial strains were cultured on PDA (potato, 200 g; glucose, 20 g;agar, 15 g; distilled water, 1 L), together with the target M. oryzae strain Y34 or JL0910 in the sameplates, and the subsequent growth inhibitions of M. oryzae were observed.

2.5. Identification of Bacterial Strains by 16S rRNA Analysis

DNA used for PCR was isolated using the method reported previously [19]. The bacterial 16SrRNA genes were amplified using forward primer 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) andreverse primer 1492R (5′-GGCTACCTTGTTACGACTT-3′) with the method reported previously [7].The purified PCR products with a size of about 1.45 kb were sequenced. The sequences were depositedin the GenBank database, and the accession numbers were MT501806 (B. safensis JLS5), MT501807(P. koreensis JLS8), MT501808 (P. saponiphila JLS10), MT501809 (S. rhizophila JLS11), and MT501810(B. tequilensis JLS12). Sequence similarity was analyzed by the BLAST search program on the websiteof the National Center for Biotechnology Information (NCBI) (https://blast.ncbi.nlm.nih.gov/Blast.cgi).For phylogenetic analysis, published reference sequences for 16S rRNA genes were obtained from theNCBI database and aligned with the sequences obtained in this study by Clustal W within MEGA X [20].

Int. J. Environ. Res. Public Health 2020, 17, 5248 3 of 12

The aligned sequences were tested for their phylogeny, and a neighbor-joining tree was generatedusing MEGA X to determine the relationships between the reference and unknown sequences.

2.6. Conidial Germination Assay

For M. oryzae conidial germination, three droplets (10 µL/each) of bacteria-treated and untreatedconidial suspensions were loaded on a sterile glass slide (3 slides/treatment) and incubated at28 ◦C in a moisture tray. The bacterial potato dextrose broth (PDB) cultures were diluted five-fold(about 108 cells/mL) and used, with the same amount of PDB as control. The conidial germination wasobserved under microscopy, and the germination rate was recorded.

2.7. Pathogenicity Suppression Assays

Rice varieties Nipponbare and Jijing88, compatible with strains Y34 and JL0910 respectively,were used for the pathogenicity assay of M. oryzae. Plants were cultivated for two weeks after sowing,at the 3- to 4-leaf stage.

Two methods were used for rice inoculation. For the drop-inoculation approach [21], rice leavesof intact seedlings with soil-free roots were spread in a container, then were inoculated by droppingM. oryzae conidial suspension (1 × 105 conidia/mL, 8 µL) on the leaf surface. Bacterial cultures wereapplied on the same point, before, simultaneously, or after M. oryzae inoculation. The spray-inoculationapproach was used for disease control assays under greenhouse conditions. Rice seedlings were sprayedwith M. oryzae conidial suspension (5 × 105 conidia/mL in 0.03% Tween 20). Then, seedlings wereplaced in a humid and dark chamber for 24 h before being incubated further in a moist atmospherewith a 12-h light period. Bacterial cultures were applied to the rice seedlings before inoculationwith M. oryzae. Symptoms were assessed visually at 7 DPI. Pathogenicity assay was performedas previously described [22]. Quantification of lesion size was performed with ImageJ software(https://imagej.nih.gov/ij/).

Detached leaves of the barley cultivar, Golden Promise, were used for pathogenicity suppressionassays. A drop-inoculation approach similar to that mentioned above was used.

3. Results

3.1. Isolation and Screening of Antagonistic Bacterial Strains against M. oryzae from Soda Saline-Sodic Rice Fields

To identify antagonistic strains for rice blast disease, we collected seven soil samples fromsoda saline-sodic rice fields in the western Songnen Plain, in the Qian Gorlos Mongol AutonomousCounty, Jilin, China. The soil samples showed pH values of 7.94–8.51, CO3

2− levels of 0.001–0.002 g/kg,and HCO3

− levels of 0.03–0.04 g/kg (Table 1).

Table 1. Soil sample parameters and the number of isolated bacterial strains.

Soil Sample pHCollection Location Altitude

(m)

Numbers of Bacteria

North Latitude East Longitude Total Antagonistic 1

#1 8.03 45◦01′06” 124◦36′10” 130 16 2#2 8.08 45◦01′06” 124◦36′10” 130 11 0#3 7.94 45◦01′06” 124◦36′10” 130 11 1#4 8.51 44◦57′19” 124◦38′14” 140 11 0#5 8.11 44◦59′40” 124◦40′54” 130 11 0#6 8.00 44◦59′40” 124◦42′47” 140 11 2#7 8.19 45◦00′00” 124◦42′27” 130 11 4

1 Antagonistic strains screened with suppression ability against M. oryzae growth.

A total of 82 strains were isolated from the seven soil samples. Next, we evaluated their inhibitionefficiency against M. oryzae strain Y34 using a two-stage screening approach. For the primary screening,up to ten bacterial strains were inoculated onto the same PDA plate, surrounding M. oryzae located at

Int. J. Environ. Res. Public Health 2020, 17, 5248 4 of 12

the center of the plate (Figure 1A). The resultant candidate strains with antagonistic phenotypes werefurther evaluated by the secondary screening, with only three bacterial strains inoculated in each plateinstead (Figure 1B). In the end, we obtained nine bacterial strains that showed significant inhibition toM. oryzae growth.

Int. J. Environ. Res. Public Health 2020, 17, x 4 of 13

M. oryzae located at the center of the plate (Figure 1A). The resultant candidate strains with

antagonistic phenotypes were further evaluated by the secondary screening, with only three bacterial

strains inoculated in each plate instead (Figure 1B). In the end, we obtained nine bacterial strains that

showed significant inhibition to M. oryzae growth.

Figure 1. Screening of antagonistic bacterial strains isolated from soda saline-alkali cultivated soils

against M. oryzae. (A, B) Primary screening (A) and secondary screening (B) are shown with typical

PDA plates. Serial numbers 4 to 15 represent candidate bacterial antagonistic strains based on primary

screening. The tested strain of M. oryzae is Y34.

3.2. Molecular Identification of the Isolated Antagonistic Bacterial Strains

Among the nine isolated antagonistic strains, five of them showed strong inhibition to M. oryzae

growth (Table 2). A phylogenetic tree based on bacterial 16S rRNA gene sequences suggests that JLS5

and JLS12 belong to Bacillus (Bacillus safensis and Bacillus tequilensis, respectively); JLS8 and JLS10

belong to Pseudomonas (Pseudomonas koreensis and Pseudomonas saponiphila, respectively); JLS11 is

Stenotrophomonas rhizophila (Figure 2 and Table 2). To our knowledge, this is the first report of

Pseudomonas saponiphila and Stenotrophomonas rhizophila with antagonistic activities against M. oryzae.

Table 2. Isolated bacterial strains with strong inhibition to M. oryzae growth.

1 Determined by plate growth inhibition. The more the number of “+” the higher the degree of inhibition.

Strain Inhibition effect on

M. oryzae growth 1 Identification Soil sample

JLS5 +++++++ Bacillus safensis #6

JLS8 + Pseudomonas koreensis #7

JLS10 +++ Pseudomonas saponiphila #7

JLS11 +++++ Stenotrophomonas rhizophila #7

JLS12 +++++++ Bacillus tequilensis #7

Figure 1. Screening of antagonistic bacterial strains isolated from soda saline-alkali cultivated soilsagainst M. oryzae. (A,B) Primary screening (A) and secondary screening (B) are shown with typicalPDA plates. Serial numbers 4 to 15 represent candidate bacterial antagonistic strains based on primaryscreening. The tested strain of M. oryzae is Y34.

3.2. Molecular Identification of the Isolated Antagonistic Bacterial Strains

Among the nine isolated antagonistic strains, five of them showed strong inhibition to M. oryzaegrowth (Table 2). A phylogenetic tree based on bacterial 16S rRNA gene sequences suggests that JLS5and JLS12 belong to Bacillus (Bacillus safensis and Bacillus tequilensis, respectively); JLS8 and JLS10belong to Pseudomonas (Pseudomonas koreensis and Pseudomonas saponiphila, respectively); JLS11 isStenotrophomonas rhizophila (Figure 2 and Table 2). To our knowledge, this is the first report ofPseudomonas saponiphila and Stenotrophomonas rhizophila with antagonistic activities against M. oryzae.

Table 2. Isolated bacterial strains with strong inhibition to M. oryzae growth.

Strain Inhibition Effect onM. oryzae Growth 1 Identification Soil Sample

JLS5 +++++++ Bacillus safensis #6JLS8 + Pseudomonas koreensis #7JLS10 +++ Pseudomonas saponiphila #7JLS11 +++++ Stenotrophomonas rhizophila #7JLS12 +++++++ Bacillus tequilensis #7

1 Determined by plate growth inhibition. The more the number of “+” the higher the degree of inhibition.

Int. J. Environ. Res. Public Health 2020, 17, 5248 5 of 12Int. J. Environ. Res. Public Health 2020, 17, x 5 of 13

Figure 2. Neighbor-joining phylogenetic tree obtained from the 16S rRNA gene sequence data of the

isolated antagonistic bacterial strains. Reference sequences obtained from the NCBI database are

indicated by their accession numbers. Numbers at nodes are percentage bootstrap values based on

1000 replicates. The isolated antagonistic bacterial strains are marked with triangles, respectively, and

their accession numbers were shown in the section of Materials and Methods.

3.3. Inhibitions of Mycelial Growth of M. Oryzae by the Isolated Antagonistic Bacterial Strains

The inhibition of M. oryzae growth by four isolated antagonistic bacterial strains was assayed

further, by culturing only one bacterial strain facing M. oryzae on each PDA plate. Two M. oryzae wild-

type strains, Y34 and JL0910 were used, respectively. The data showed that B. safensis JLS5 and B.

tequilensis JLS12 could greatly inhibit the growth of M. oryzae Y34, with the growth rates reduced to

about 20% of the control, i.e., showing 80% inhibition against Y34. The strain S. rhizophila JLS11

showed growth inhibition of about 70%. The suppression of M. oryzae growth by P. saponiphila JLS10

was relatively weak, with about 45% inhibition (Figure 3A,B,D,E). Similar results were also observed

for the M. oryzae strain, JL0910 (Figure 3C–E). Our data showed that B. safensis JLS5 and B. tequilensis

JLS12 exhibited the most effective biocontrol potential; therefore, both strains were studied further

for the control of rice blast disease.

Figure 2. Neighbor-joining phylogenetic tree obtained from the 16S rRNA gene sequence data ofthe isolated antagonistic bacterial strains. Reference sequences obtained from the NCBI database areindicated by their accession numbers. Numbers at nodes are percentage bootstrap values based on1000 replicates. The isolated antagonistic bacterial strains are marked with triangles, respectively,and their accession numbers were shown in the section of Materials and Methods.

3.3. Inhibitions of Mycelial Growth of M. Oryzae by the Isolated Antagonistic Bacterial Strains

The inhibition of M. oryzae growth by four isolated antagonistic bacterial strains was assayedfurther, by culturing only one bacterial strain facing M. oryzae on each PDA plate. Two M. oryzaewild-type strains, Y34 and JL0910 were used, respectively. The data showed that B. safensis JLS5 andB. tequilensis JLS12 could greatly inhibit the growth of M. oryzae Y34, with the growth rates reducedto about 20% of the control, i.e., showing 80% inhibition against Y34. The strain S. rhizophila JLS11showed growth inhibition of about 70%. The suppression of M. oryzae growth by P. saponiphila JLS10was relatively weak, with about 45% inhibition (Figure 3A,B,D,E). Similar results were also observedfor the M. oryzae strain, JL0910 (Figure 3C–E). Our data showed that B. safensis JLS5 and B. tequilensisJLS12 exhibited the most effective biocontrol potential; therefore, both strains were studied further forthe control of rice blast disease.

Int. J. Environ. Res. Public Health 2020, 17, 5248 6 of 12

Int. J. Environ. Res. Public Health 2020, 17, x 6 of 13

Figure 3. Inhibitions of M. oryzae growth by the isolated bacterial strains. (A) Colony morphology of

four candidate biocontrol bacterial strains on LB plates. (B) Inhibitions of the colony growth of M.

oryzae (Mo) strain Y34 by the four isolated bacterial strains. Representative photographs were taken

at 10 days post-inoculation (DPI). S5, S10, S11, and S12 indicates bacterial strain JLS5, JLS10, JLS11,

and JLS12, respectively. (C) Inhibition of the colony growth of Mo strain JL0910 by the four bacterial

strains. Representative photographs were taken at 7 DPI. (D) Growth of the two M. oryzae strains, as

control (CK). (E) Quantitative analysis of the growth inhibition caused by the indicated strains in (B)

and (C). Data represent means standard deviation (SD) from at least three independent experiments.

** indicates statistical significance at p < 0.01.

Figure 3. Inhibitions of M. oryzae growth by the isolated bacterial strains. (A) Colony morphologyof four candidate biocontrol bacterial strains on LB plates. (B) Inhibitions of the colony growth ofM. oryzae (Mo) strain Y34 by the four isolated bacterial strains. Representative photographs were takenat 10 days post-inoculation (DPI). S5, S10, S11, and S12 indicates bacterial strain JLS5, JLS10, JLS11,and JLS12, respectively. (C) Inhibition of the colony growth of Mo strain JL0910 by the four bacterialstrains. Representative photographs were taken at 7 DPI. (D) Growth of the two M. oryzae strains,as control (CK). (E) Quantitative analysis of the growth inhibition caused by the indicated strains in(B,C). Data represent means ± standard deviation (SD) from at least three independent experiments.** indicates statistical significance at p < 0.01.

3.4. Bacterial Strains JLS5 and JLS12 Inhibit Conidial Germination of M. Oryzae

Conidium is the primary source for M. oryzae spreading and infection in the field; inhibition ofconidial germination is vital for rice blast disease control. Therefore, we evaluated the suppressionefficiency of conidial germination by the isolated antagonistic bacterial strains JLS5 and JLS12,using five-fold dilutions of their PDB cultures (about 108 cells/mL in the final content). The resultsindicated that both JLS5 and JLS12 could inhibit conidial germination of M. oryzae effectively.

Int. J. Environ. Res. Public Health 2020, 17, 5248 7 of 12

JLS5 treatment could inhibit 95% of conidial germination, showing only 5% of conidia germinatedat 12 h post-incubation, compared to over 90% conidial germination in the control (Figure 4).JLS12 treatment could inhibit 85% of conidial germination. A similar result was also observedfor the treatment with cell-free culture filtrates of JLS12, which inhibited about 81% of conidialgermination (Figure 4). Notably, many bacteria-treated conidia suffered morphological changes orrupture. The resultant germinated conidia grew very slowly, with most of them showing very shortgerm tubes compared to that of the control (Figure 4).

Int. J. Environ. Res. Public Health 2020, 17, x 7 of 13

3.4. Bacterial Strains JLS5 and JLS12 Inhibit Conidial Germination of M. Oryzae

Conidium is the primary source for M. oryzae spreading and infection in the field; inhibition of

conidial germination is vital for rice blast disease control. Therefore, we evaluated the suppression

efficiency of conidial germination by the isolated antagonistic bacterial strains JLS5 and JLS12, using

five-fold dilutions of their PDB cultures (about 108 cells/mL in the final content). The results indicated

that both JLS5 and JLS12 could inhibit conidial germination of M. oryzae effectively. JLS5 treatment

could inhibit 95% of conidial germination, showing only 5% of conidia germinated at 12 h post-

incubation, compared to over 90% conidial germination in the control (Figure 4). JLS12 treatment

could inhibit 85% of conidial germination. A similar result was also observed for the treatment with

cell-free culture filtrates of JLS12, which inhibited about 81% of conidial germination (Figure 4).

Notably, many bacteria-treated conidia suffered morphological changes or rupture. The resultant

germinated conidia grew very slowly, with most of them showing very short germ tubes compared

to that of the control (Figure 4).

Figure 4. Inhibition of M. oryzae conidial germination by JLS5 or JLS12. (A) Conidial germination of

M. oryzae Y34 is inhibited by the indicated bacterial strains, respectively. Scale bar indicates 50 μM.

JLS12-F indicates the cell-free culture filtrates of JLS12. Representative photographs were taken after

12 h incubation. (B) Quantitative analysis of the conidial germination inhibition caused by the

indicated strains in (A). Data represent means SD from at least three independent experiments. ***

indicates statistical significance at p < 0.001.

3.5. Suppressions of M. Oryzae Pathogenicity on Host Leaves by the Isolated Antagonistic Bacterial Strains

JLS5 and JLS12

To evaluate the effects of JLS5 and JLS12 on M. oryzae pathogenicity, we inoculated their PDB

cultures on detached barley leaves together with M. oryzae, and before or after M. oryzae inoculation

at different time points (Figure 5A). Applications of bacterial culture before M. oryzae inoculation are

to mimic preventive treatment, and applications after M. oryzae inoculation are for curative test. The

application amount for each bacterial strain is a five-fold dilution of PDB culture (about 108 cells/mL),

the same as that involved in the conidial germination assay.

Figure 4. Inhibition of M. oryzae conidial germination by JLS5 or JLS12. (A) Conidial germinationof M. oryzae Y34 is inhibited by the indicated bacterial strains, respectively. Scale bar indicates50 µM. JLS12-F indicates the cell-free culture filtrates of JLS12. Representative photographs were takenafter 12 h incubation. (B) Quantitative analysis of the conidial germination inhibition caused by theindicated strains in (A). Data represent means ± SD from at least three independent experiments.*** indicates statistical significance at p < 0.001.

3.5. Suppressions of M. Oryzae Pathogenicity on Host Leaves by the Isolated Antagonistic Bacterial StrainsJLS5 and JLS12

To evaluate the effects of JLS5 and JLS12 on M. oryzae pathogenicity, we inoculated theirPDB cultures on detached barley leaves together with M. oryzae, and before or after M. oryzaeinoculation at different time points (Figure 5A). Applications of bacterial culture before M. oryzaeinoculation are to mimic preventive treatment, and applications after M. oryzae inoculation are forcurative test. The application amount for each bacterial strain is a five-fold dilution of PDB culture(about 108 cells/mL), the same as that involved in the conidial germination assay.

The results showed that both bacterial strains shared similar inhibition effects on M. oryzaepathogenicity, and all treatments could inhibit the pathogenicity significantly (Figure 5). The mosteffective control period for bacterial culture application is from 24 HBI to 12 HPI, during whichthe blast fungus has not penetrated into the host cells yet (Figure 5B). Applications of bacterialculture in this period could effectively inhibit the severity of rice blast, with suppression over 80%.Bacterial application at 24 HPI, when the blast fungus started to penetrate and colonize host cells,decreased the suppression ability to about 70%; when applying at 48 HPI, the inhibition was about 30%(Figure 5C). Similar results were observed when cell-free culture filtrates of JLS12 were used, and allapplications from 24 HBI to 48 HPI could inhibit M. oryzae pathogenicity significantly (Figure 5B,C).

Int. J. Environ. Res. Public Health 2020, 17, 5248 8 of 12

Int. J. Environ. Res. Public Health 2020, 17, x 8 of 13

Figure 5. Suppression of M. oryzae pathogenicity on barley leaves by JLS5 or JLS12. (A) Schematic

diagram of application time of biocontrol bacteria. According to the inoculation time of M. oryzae,

biocontrol bacteria were applied at 24 h before inoculation (HBI), 0 (applied with M. oryzae conidia

simultaneously), 5 h post inoculation (HPI), 12 HPI, 24 HPI and 48 HPI, respectively. (B) The

pathogenicity of M. oryzae Y34 is inhibited by the indicated bacterial strains, respectively. M. oryzae

were inoculated with a conidial drop on detached barley leaves. JLS12-F indicates the cell-free culture

filtrates of JLS12. Representative photographs were taken at 7 days post-inoculation. (C) Quantitative

analysis of the pathogenicity inhibition caused by the indicated strains in (B). −24, 0, 5, 12, 24, and 48

indicates 24 HBI, 0, 5 HPI, 12 HPI, 24 HPI, and 48 HPI, respectively. Data represent means SD from

at least three independent experiments. ** indicates statistical significance at p < 0.01.

Figure 5. Suppression of M. oryzae pathogenicity on barley leaves by JLS5 or JLS12. (A) Schematicdiagram of application time of biocontrol bacteria. According to the inoculation time of M. oryzae,biocontrol bacteria were applied at 24 h before inoculation (HBI), 0 (applied with M. oryzae conidiasimultaneously), 5 h post inoculation (HPI), 12 HPI, 24 HPI and 48 HPI, respectively. (B) Thepathogenicity of M. oryzae Y34 is inhibited by the indicated bacterial strains, respectively. M. oryzae wereinoculated with a conidial drop on detached barley leaves. JLS12-F indicates the cell-free culture filtratesof JLS12. Representative photographs were taken at 7 days post-inoculation. (C) Quantitative analysisof the pathogenicity inhibition caused by the indicated strains in (B). −24, 0, 5, 12, 24, and 48 indicates24 HBI, 0, 5 HPI, 12 HPI, 24 HPI, and 48 HPI, respectively. Data represent means ± SD from at leastthree independent experiments. ** indicates statistical significance at p < 0.01.

Int. J. Environ. Res. Public Health 2020, 17, 5248 9 of 12

To confirm the pathogenicity inhibition observed on barley, we did similar tests on rice withthe drop-inoculation method (Figure 6A), and similar results were observed. Applications of JLS5 orJLS12, during the period from 12 HBI to 12 HPI, inhibited M. oryzae pathogenicity effectively, with theinhibition over 90% and 80%, respectively (Figure 6B,C). Even application at 24 HPI, the inhibitionremained at 65% and 45%, respectively (Figure 6B,C).

Int. J. Environ. Res. Public Health 2020, 17, x 9 of 13

The results showed that both bacterial strains shared similar inhibition effects on M. oryzae

pathogenicity, and all treatments could inhibit the pathogenicity significantly (Figure 5). The most

effective control period for bacterial culture application is from 24 HBI to 12 HPI, during which the

blast fungus has not penetrated into the host cells yet (Figure 5B). Applications of bacterial culture in

this period could effectively inhibit the severity of rice blast, with suppression over 80%. Bacterial

application at 24 HPI, when the blast fungus started to penetrate and colonize host cells, decreased

the suppression ability to about 70%; when applying at 48 HPI, the inhibition was about 30% (Figure

5C). Similar results were observed when cell-free culture filtrates of JLS12 were used, and all

applications from 24 HBI to 48 HPI could inhibit M. oryzae pathogenicity significantly (Figure 5B,C).

To confirm the pathogenicity inhibition observed on barley, we did similar tests on rice with the

drop-inoculation method (Figure 6A), and similar results were observed. Applications of JLS5 or

JLS12, during the period from 12 HBI to 12 HPI, inhibited M. oryzae pathogenicity effectively, with

the inhibition over 90% and 80%, respectively (Figure 6B,C). Even application at 24 HPI, the inhibition

remained at 65% and 45%, respectively (Figure 6B,C).

Figure 6. Suppression of M. oryzae pathogenicity on rice leaves by JLS5 or JLS12. (A) Schematic

diagram of application time of biocontrol bacteria. According to the inoculation time of M. oryzae,

biocontrol bacteria were applied at 12 h before inoculation (HBI), 0 (applied with M. oryzae conidia

simultaneously), 5 h post-inoculation (HPI), 12 HPI and 24 HPI, respectively. (B) The pathogenicity

of M. oryzae Y34 is inhibited by the indicated bacterial strains, respectively. M. oryzae were inoculated

with a conidial drop on rice leaves. JLS12-F indicates the cell-free culture filtrates of JLS12.

Representative photographs were taken at 7 days post-inoculation. (C) Quantitative analysis of the

pathogenicity inhibition caused by the indicated strains in (B). −12, 0, 5, 12 and 24 indicates 12 HBI, 0,

5 HPI, 12 HPI and 24 HPI, respectively. Data represent means SD from at least three independent

experiments. ** indicates statistical significance at p < 0.01.

These tests indicated that both JLS5 and JLS12 display excellent preventive effects on the

pathogenicity of M. oryzae. Although a good curative effect can be obtained when applied at the early

infectious stage of the rice blast fungus, the control effect of the bacterial strains decreased

significantly. Therefore, when using these biocontrol bacteria to control rice blast disease, prevention

should be the priority.

Figure 6. Suppression of M. oryzae pathogenicity on rice leaves by JLS5 or JLS12. (A) Schematicdiagram of application time of biocontrol bacteria. According to the inoculation time of M. oryzae,biocontrol bacteria were applied at 12 h before inoculation (HBI), 0 (applied with M. oryzaeconidia simultaneously), 5 h post-inoculation (HPI), 12 HPI and 24 HPI, respectively. (B) Thepathogenicity of M. oryzae Y34 is inhibited by the indicated bacterial strains, respectively. M. oryzae wereinoculated with a conidial drop on rice leaves. JLS12-F indicates the cell-free culture filtrates of JLS12.Representative photographs were taken at 7 days post-inoculation. (C) Quantitative analysis of thepathogenicity inhibition caused by the indicated strains in (B). −12, 0, 5, 12 and 24 indicates 12 HBI,0, 5 HPI, 12 HPI and 24 HPI, respectively. Data represent means ± SD from at least three independentexperiments. ** indicates statistical significance at p < 0.01.

These tests indicated that both JLS5 and JLS12 display excellent preventive effects on thepathogenicity of M. oryzae. Although a good curative effect can be obtained when applied at the earlyinfectious stage of the rice blast fungus, the control effect of the bacterial strains decreased significantly.Therefore, when using these biocontrol bacteria to control rice blast disease, prevention should bethe priority.

3.6. Control of Rice Blast Disease on Rice Seedlings by the Isolated Antagonistic Bacterial Strains JLS5 andJLS12 in Greenhouse

Since JLS5 and JLS12 exhibited strong preventive suppressions against M. oryzae, we subsequentlyevaluated their control effects on rice blast disease in the greenhouse with a preventive approach.Rice seedlings were sprayed with five-fold dilutions of bacterial PDB cultures with 108 cells/mL incontent at 36 h before inoculation of M. oryzae (Figure 7A). The results showed that rice blast diseasewas controlled effectively by the preventive application of JLS5 and JLS12, respectively. Only rare tinylesions appeared on rice leaves, which were pre-applied with bacterial cultures (Figure 7B); over 85% ofinhibition was observed based on lesion size calculation (Figure 7C). A similar result was found when

Int. J. Environ. Res. Public Health 2020, 17, 5248 10 of 12

20% of cell-free culture filtrates of JLS12 were used (Figure 7B,C). These greenhouse tests indicated thatthe PDB cultures of JLS5 and JLS12 could be used for effective control of rice blast disease.

Int. J. Environ. Res. Public Health 2020, 17, x 10 of 13

3.6. Control of Rice Blast Disease on Rice Seedlings by the Isolated Antagonistic Bacterial Strains JLS5 and

JLS12 in Greenhouse

Since JLS5 and JLS12 exhibited strong preventive suppressions against M. oryzae, we

subsequently evaluated their control effects on rice blast disease in the greenhouse with a preventive

approach. Rice seedlings were sprayed with five-fold dilutions of bacterial PDB cultures with 108

cells/mL in content at 36 h before inoculation of M. oryzae (Figure 7A). The results showed that rice

blast disease was controlled effectively by the preventive application of JLS5 and JLS12, respectively.

Only rare tiny lesions appeared on rice leaves, which were pre-applied with bacterial cultures (Figure

7B); over 85% of inhibition was observed based on lesion size calculation (Figure 7C). A similar result

was found when 20% of cell-free culture filtrates of JLS12 were used (Figure 7B,C). These greenhouse

tests indicated that the PDB cultures of JLS5 and JLS12 could be used for effective control of rice blast

disease.

Figure 7. Control of rice blast by JLS5 or JLS12 under greenhouse conditions. (A) Schematic diagram

of the treatment time of biocontrol bacteria. Bacterial strains were applied at 36 h before inoculation

(HBI) of M. oryzae. (B) The pathogenicity of M. oryzae Y34 is inhibited by the indicated bacterial strains,

respectively. M. oryzae were inoculated by conidial spraying on rice seedlings in the greenhouse.

JLS12-F indicates the cell-free culture filtrates of JLS12. Representative photographs were taken at 7

days post-inoculation. (C) Quantitative analysis of the pathogenicity inhibition caused by the

indicated strains in (B). Data represent means SD from at least three independent experiments. ***

indicates statistical significance at p < 0.001.

4. Discussion

In recent years, quite a few potential biocontrol strains for rice blast suppression have been

isolated from plant rhizosphere soils [9–13]; however, none of them were collected from saline-sodic

fields. In this study, we have isolated 82 bacterial strains from the soda saline-sodic fields in the

western Songnen Plain of China, with several of the strains showing potentials as biological control

agents of M. oryzae. Saline-sodic soil is a type of degraded land, and approximately 1.1 × 109 hectares

of soil worldwide are saline-affected [23]. The western Songnen Plain is one of the three major regions

with saline-sodic soils in the world containing over 3.0 × 106 hectares of soda saline-sodic lands [24].

Our study indicates that these lands are rich sources for biocontrol strain isolation to inhibit rice blast

or other plant fungal diseases. To our knowledge, this is the first report of biocontrol strain isolation

from soda saline-sodic fields.

We have identified five bacterial strains with significant inhibition against M. oryzae in this

study, which are B. safensis JLS5, P. koreensis JLS8, P. saponiphila JLS10, S. rhizophila JLS11 and B.

tequilensis JLS12. To our knowledge, this is the first report showing that P. saponiphila and S. rhizophila

display antagonistic activities against M. oryzae; their applications in controlling rice blast and the

associated mechanisms should be studied further. The five antagonistic bacterial strains isolated in

this study do not show adverse effects on rice growth and development. Among them, B. safensis JLS5

and B. tequilensis JLS12 exhibited strong suppression of mycelial growth, conidial germination, and

Figure 7. Control of rice blast by JLS5 or JLS12 under greenhouse conditions. (A) Schematic diagramof the treatment time of biocontrol bacteria. Bacterial strains were applied at 36 h before inoculation(HBI) of M. oryzae. (B) The pathogenicity of M. oryzae Y34 is inhibited by the indicated bacterialstrains, respectively. M. oryzae were inoculated by conidial spraying on rice seedlings in the greenhouse.JLS12-F indicates the cell-free culture filtrates of JLS12. Representative photographs were taken at 7 dayspost-inoculation. (C) Quantitative analysis of the pathogenicity inhibition caused by the indicated strainsin (B). Data represent means ± SD from at least three independent experiments. *** indicates statisticalsignificance at p < 0.001.

4. Discussion

In recent years, quite a few potential biocontrol strains for rice blast suppression have been isolatedfrom plant rhizosphere soils [9–13]; however, none of them were collected from saline-sodic fields.In this study, we have isolated 82 bacterial strains from the soda saline-sodic fields in the westernSongnen Plain of China, with several of the strains showing potentials as biological control agentsof M. oryzae. Saline-sodic soil is a type of degraded land, and approximately 1.1 × 109 hectares ofsoil worldwide are saline-affected [23]. The western Songnen Plain is one of the three major regionswith saline-sodic soils in the world containing over 3.0 × 106 hectares of soda saline-sodic lands [24].Our study indicates that these lands are rich sources for biocontrol strain isolation to inhibit rice blastor other plant fungal diseases. To our knowledge, this is the first report of biocontrol strain isolationfrom soda saline-sodic fields.

We have identified five bacterial strains with significant inhibition against M. oryzae in this study,which are B. safensis JLS5, P. koreensis JLS8, P. saponiphila JLS10, S. rhizophila JLS11 and B. tequilensis JLS12.To our knowledge, this is the first report showing that P. saponiphila and S. rhizophila display antagonisticactivities against M. oryzae; their applications in controlling rice blast and the associated mechanismsshould be studied further. The five antagonistic bacterial strains isolated in this study do not showadverse effects on rice growth and development. Among them, B. safensis JLS5 and B. tequilensis JLS12exhibited strong suppression of mycelial growth, conidial germination, and pathogenicity of the riceblast fungus. This inhibition was much more effective when the bacterial agents were applied beforeM. oryzae penetration. The results suggest that, when using these strains to manage rice blast disease,we should apply them before symptoms appear, which is also the common feature of biological control.Recently, we have been trying to apply B. tequilensis JLS12 to manage rice blast disease in the field withpreventive strategy. The control effect of the first-stage application appeared encouraging, which willbe confirmed further in the subsequent growing seasons.

It should be pointed out that two related strains, B. safensis B21 [8] and B. tequilensis GYLH001 [7],have been reported previously as potential biological control agents of M. oryzae. Both strains were also

Int. J. Environ. Res. Public Health 2020, 17, 5248 11 of 12

isolated from China; however, in contrast to our strains (B. safensis JLS5 and B. tequilensis JLS12) thatare isolated in cold winter from soda saline-sodic soils located in Northeast China, both of the otherreported strains GYLH001 and B21 were plant endophytic bacteria isolated in the growing season fromAngelica dahurica (a traditional Chinese medicine) and Osmanthus fragrans located in warmer regionsof China, respectively. Therefore, we suspect that these strains would show distinct features that arerepresentative of their ecological niches. The difference in biological control mechanisms between themneeds to be studied further. Since the bacterial strains reported here were isolated from extreme coldfrozen soil (below −20 ◦C) in winter, they are suitable for protecting rice production in cold regions(e.g., Northeast China). In addition to being effective agents for rice blast management in growingseasons, these cold-adapted bacterial strains have the potential to inhibit M. oryzae (conidia and/ormycelia) survival in diseased rice tissues or soils during the non-growing season, and to attenuatetheir primary infection in the next growing season. This non-growing season management of rice blastis usually ignored in the current popular biological control strategies. The promising application ofinhibiting the survival of the blast fungus in non-growing seasons is worthy of a comprehensive studyin the near future.

In summary, five bacterial strains with significant inhibition against M. oryzae growth wereisolated from soda saline-sodic soils in this study. Two of them showed strong suppression of conidialgermination and pathogenicity of M. oryzae, indicating the potential application of both strains asbiological control agents of rice blast. Our data suggests that soda saline-sodic soils can be rich sourcesfor biocontrol strain isolation.

5. Conclusions

Our study is the first report to successfully isolate and identify biocontrol bacterial strains fromsoda saline-sodic land for the management of rice blast disease. A total of 82 bacterial strains wereisolated, and five of them exhibited strong inhibition of mycelial growth of M. oryzae. The strains wereidentified as B. safensis JLS5, P. koreensis JLS8, P. saponiphila JLS10, S. rhizophila JLS11, and B. tequilensisJLS12. PDB cultures of JLS5 or JLS12 exhibited strong inhibition effects on conidial germinationand pathogenicity of the rice blast fungus, both preventively and curatively. The suppressions ofpathogenicity were further confirmed by greenhouse experiments, indicating that both strains havepotential applications for biological control of M. oryzae. This study suggests that soda saline-sodicsoils can be rich sources for biocontrol strain isolation.

Author Contributions: Conceptualization, G.L.; methodology, Y.W., L.L., W.H., H.J., M.Z., and G.L.; resources,H.J., S.Z., and G.L.; supervision, G.L., Q.Q., and S.Z.; writing—original draft preparation, G.L. All authors haveread and agreed to the published version of the manuscript.

Funding: This research was supported by the National Key R&D Program of China (grant number2018YFD0200202), the grant of the China Scholarship Council (grant number 201906175076) and the NationalUndergraduate Innovative and Entrepreneurship Training Program (201910183016).

Acknowledgments: We thank Xin Wang from Miami University for kindly helping grammar revision. The authorswould like to thank the anonymous reviewers for their constructive comments on this article.

Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of thestudy; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision topublish the results.

References

1. Howard, R.J.; Valent, B. Breaking and entering: Host penetration by the fungal rice blast pathogenMagnaporthe grisea. Annu. Rev. Microbiol. 1996, 50, 491–512. [CrossRef] [PubMed]

2. Talbot, N.J. On the trail of a cereal killer: Exploring the biology of Magnaporthe grisea. Annu. Rev. Microbiol.2003, 57, 177–202. [CrossRef] [PubMed]

3. Deng, Y.Z.; Naqvi, N.I. Metabolic Basis of Pathogenesis and Host Adaptation in Rice Blast. Annu. Rev. Microbiol.2019, 73, 601–619. [CrossRef] [PubMed]

Int. J. Environ. Res. Public Health 2020, 17, 5248 12 of 12

4. Leneveu-Jenvrin, C.; Charles, F.; Barba, F.J.; Remize, F. Role of biological control agents and physicaltreatments in maintaining the quality of fresh and minimally-processed fruit and vegetables. Crit. Rev. FoodSci. Nutr. 2019, 60, 1–19. [CrossRef] [PubMed]

5. Moraes Bazioli, J.; Belinato, J.R.; Costa, J.H.; Akiyama, D.Y.; Pontes, J.G.M.; Kupper, K.C.; Augusto, F.;de Carvalho, J.E.; Fill, T.P. Biological Control of Citrus Postharvest Phytopathogens. Toxins 2019, 11, 460.[CrossRef] [PubMed]

6. Kohl, J.; Kolnaar, R.; Ravensberg, W.J. Mode of Action of Microbial Biological Control Agents Against PlantDiseases: Relevance Beyond Efficacy. Front. Plant. Sci. 2019, 10, 845. [CrossRef]

7. Li, H.; Guan, Y.; Dong, Y.; Zhao, L.; Rong, S.; Chen, W.; Lv, M.; Xu, H.; Gao, X.; Chen, R.; et al. Isolation andevaluation of endophytic Bacillus tequilensis GYLH001 with potential application for biological control ofMagnaporthe oryzae. PLoS ONE 2018, 13, e0203505. [CrossRef]

8. Rong, S.; Xu, H.; Li, L.; Chen, R.; Gao, X.; Xu, Z. Antifungal activity of endophytic Bacillus safensis B21 and itspotential application as a biopesticide to control rice blast. Pestic Biochem. Physiol. 2020, 162, 69–77. [CrossRef]

9. Filippi, M.C.C.; da Silva, G.B.; Silva-Lobo, V.L.; Côrtes, M.V.C.B.; Moraes, A.J.G.; Prabhu, A.S. Leaf blast(Magnaporthe oryzae) suppression and growth promotion by rhizobacteria on aerobic rice in Brazil. Biol. Control.2011, 58, 160–166. [CrossRef]

10. Huang, W.; Liu, X.; Zhou, X.; Wang, X.; Liu, X.; Liu, H. Calcium Signaling Is Suppressed in Magnaporthe oryzaeConidia by Bacillus cereus HS24. Phytopathology 2020, 110, 309–316. [CrossRef]

11. Tokpah, D.P.; Li, H.; Wang, L.; Liu, X.; Mulbah, Q.S.; Liu, H. An assessment system for screening effectivebacteria as biological control agents against Magnaporthe grisea on rice. Biol. Control. 2016, 103, 21–29. [CrossRef]

12. Chaiharn, M.; Theantana, T.; Pathom-Aree, W. Evaluation of Biocontrol Activities of Streptomyces spp. againstRice Blast Disease Fungi. Pathogens 2020, 9, 126. [CrossRef]

13. Yu, W.Q.; Zheng, G.P.; Qiu, W.; Yan, F.C.; Liu, W.Z.; Liu, W.X. Draft genome sequence, disease-resistancegenes, and phenotype of a Paenibacillus terrae strain (NK3-4) with the potential to control plant diseases.Genome 2018, 61, 725–734. [CrossRef] [PubMed]

14. Yu, Q.; Liu, Z.; Lin, D.; Zhang, W.; Sun, Q.; Zhu, J.; Lin, M. Characterization and evaluation of Staphylococcus sp.strain LZ16 for the biological control of rice blast caused by Magnaporthe oryzae. Biol. Control. 2013, 65, 338–347.[CrossRef]

15. Li, Q.; Jiang, Y.; Ning, P.; Zheng, L.; Huang, J.; Li, G.; Jiang, D.; Hsiang, T. Suppression of Magnaporthe oryzaeby culture filtrates of Streptomyces globisporus JK-1. Biol. Control. 2011, 58, 139–148. [CrossRef]

16. Zheng, F.; Yang, Q.; Zhao, Z.; Li, J. Variability of pathogenicity of Pyricularia oryzae. J. Yunnan Agric. Univ.1998, 13, 20–24.

17. Li, G.; Zhou, Z.; Liu, G.; Zheng, F.; He, C. Characterization of T-DNA insertion patterns in the genome of riceblast fungus Magnaporthe oryzae. Curr. Genet. 2007, 51, 233–243. [CrossRef]

18. Cui, X.; Wei, Y.; Wang, Y.H.; Li, J.; Wong, F.L.; Zheng, Y.J.; Yan, H.; Liu, S.S.; Liu, J.L.; Jia, B.L.; et al. Proteinsinteracting with mitochondrial ATP-dependent Lon protease (MAP1) in Magnaporthe oryzae are involved inrice blast disease. Mol. Plant. Pathol. 2015, 16, 847–859. [CrossRef]

19. Wilson, K. Preparation of genomic DNA from bacteria. Curr. Protoc. Mol. Biol. 2001, 56, 2–4. [CrossRef]20. Kumar, S.; Stecher, G.; Li, M.; Knyaz, C.; Tamura, K. MEGA X: Molecular Evolutionary Genetics Analysis

across Computing Platforms. Mol. Biol. Evol. 2018, 35, 1547–1549. [CrossRef]21. Koga, H.; Dohi, K.; Nakayachi, O.; Mori, M. A novel inoculation method of Magnaporthe grisea for cytological

observation of the infection process using intact leaf sheaths of rice plants. Physiol. Mol. Plant. Pathol. 2004,64, 67–72. [CrossRef]

22. Long, B.; Li, G.-H.; Feng, R.-R.; Tao, L.; Liu, D.-D.; Feng, H.-Q.; Qin, Q.-M. NTPDase Specific InhibitorsSuppress Rice Infection by Magnaporthe oryzae. J. Phytopathol. 2015, 163, 227–232. [CrossRef]

23. Stille, L.; Smeets, E.; Wicke, B.; Singh, R.; Singh, G. The economic performance of four (agro-) forestry systemson alkaline soils in the state of Haryana in India. Energy Sustain. Dev. 2011, 15, 388–397. [CrossRef]

24. Yang, F.; An, F.; Ma, H.; Wang, Z.; Zhou, X.; Liu, Z. Variations on Soil Salinity and Sodicity and Its DrivingFactors Analysis under Microtopography in Different Hydrological Conditions. Water 2016, 8, 227. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open accessarticle distributed under the terms and conditions of the Creative Commons Attribution(CC BY) license (http://creativecommons.org/licenses/by/4.0/).